|| Checking for direct PDF access through Ovid
Transgene flow from genetically engineered (GE) crops to non-GE varieties and wild relatives via pollen-mediated gene flow (PMGF) may create food and environmental biosafety concerns. Assessing the level of PMGF from GE crops is required before commercialization. Whether the level of PMGF estimated at relatively small scales can sufficiently represent the actual scenario at large production scales remains unresolved. Here, we estimated average PMGF frequencies from three insect-resistant GE rice lines to their non-GE counterparts at four scales ranging from 9 to 576 m2, having the number of GE to non-GE plants constantly at the ratio of 8:1. Based on nearly 1.3 million examined seedlings from non-GE rice plots, very low frequencies (<0.1%) of transgene flow were detected. The highest frequencies were found in plots at the smallest scales. Scale had a significantly negative effect on the frequency of PMGF in rice, with decreased gene flow at increased scale. An extended PMGF model could well represent the experimental data. Field experiments at relatively small scales combined with mathematical modelling could provide useful prediction on the level of rice transgene flow at large production scales. This is probably applicable for other crop species with wind- and self-pollination.